Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "Discovery of a close companion to the young star Haro 6-37"

See other formats








A&A manuscript no. 

(will be inserted by hand later] 

Your thesaurus codes are: 

06(03.20.2; 05.15.1; 08.02.1; 08.09.2: Haro 6-37; 08.16.5; 13.09.6) 



Discovery of a close companion to the young star Haro 6-37^ 

A. Richichi^, R. Kohler^, J. Woitas^, and Ch. Leinert^ 

^ Osservatorio Astrofisico di Arcetri, L.go Enrico Fermi 5, 1-50125 Firenze, Italy 

Astrophysikalisches Institut Potsdam, An der Sternwarte 16, D-14482 Potsdam, Germany 
^ Max-PIanck-Institut fur Astronomie, Konigstuhl 17, D-69117 Heidelberg, Germany 

' Received December 7, 1998 - Accepted January 22, 1999 

Abstract. We report lunar occultation (LO) and speckle 
interferometry observations of Haro 6-37, all carried out at 
A=2.2/Lini. The main finding is the discovery of a new close 
companion by LO, subsequently confirmed by speckle. The 
two techniques yield values in excellent agreement, with 
0'.'331 and 181° for the separation and position angle re- 
spectively. Additionally, the LO data show evidence of 
what might be interpreted as scattered light from the in- 
ner part of the dust disk which is known to exist around 
Haro 6-37 from mm-radio observations (Osterloh & Beck- 
with ( 1995 )). However this last conclusion is subject to 
some uncertainty because of possible contamination of the 
signal during data acquisition. LO and speckle results are 
in good agreement also for what concerns the brightness 
ratio of the newly found companion to the main compo- 
nent, R=0.10 or AK=2.5. We discuss the possible implica- 
tions on the general picture of Haro 6-37, in particular for 
what concerns the ages of the components and the dust 

Key words: Techniques: interferometric - Occultations 
- Stars: binaries - Stars: individual: Haro 6-37 - Stars: 
pre-main sequence - Infrared: stars 

1. Introduction 

Pre-main sequence (PMS) stars represent the link between 
the early stages of star formation, and the beginning of life 
as a normal star on the main sequence. This phase, last- 
ing « 10^ — 10^ years, is strongly influenced by the initial 
conditions of the surrounding environment, and charac- 
terizes in a significant way the newborn stars in the re- 
maining w 10® — lO^*' years of their lifetimes. Therefore, 
observations of PMS stars play an important role in our 
understanding of the processes of star formation and stel- 
lar evolution, and as such they have been studied with 

Send offprint requests to: A. Richichi 
e-mail; richichiOarcetri . astro . it 

* Based on observations collected at the Calar Alto Observa- 
tory (Spain), operated by the German-Spanish Astronomical 

increasing interest at several wavelengths from the X-rays 
to the radio domains. 

In particular, in the past decade it has become increas- 
ingly evident that many PMS stars exhibit two important 
features, namely the presence of circumstellar disks and 
a binary (or multiple) nature. The circumstellar disks, 
first inferred from millimeter observations and later ob- 
served in some cases also by direct imaging at visual/IR 
wavelengths, are thought to be the remnant of the orig- 
inal parent condensation. The presence of one or more 
physically associated companions, on the other hand, is 
the signature of a common formation process and pro- 
vides a key to discriminate between the different possible 
formation mechanism at play. The study of such binary 
or multiple systems also allows us to make inferences on 
their evolution in the presence of the complex environ- 
ment of the parent star forming region (SFR), and once 
they reach the main sequence. Of particular interest is the 
understanding of the interactions that take place between 
the circumstellar (or sometimes circumbinary) disks, and 
the binary components. This in turn has direct implica- 
tions on the formation of planetary systems, another topic 
which is receiving the highest attention in modern obser- 
vational astronomy. Useful references to many topics re- 
lated to PMS stars research can by found in the work by 

Hartmann (1998); recent compilations of PMS multiple 

stars are given by Ghez (1998) and by Mathieu (1994) 

In this line of research, high angular resolution obser- 
vations in the near-infrared (NIK) play an important role 
next to the radio and visual ranges. In fact, most radio 
observations lack the resolution to study the hot, inner 
parts of the circumstellar disks and to reveal the closest 
binary components. Observations in the visual range, on 
the other hand, are severely limited by extinction due to 
both the circumstellar dust, and the local dust in the sur- 
rounding SFR. The typical size of circumstellar disks and 
binary separations (10-100 AU) are such, that even in the 
nearest SFRs with a high number of PMS stars (Taurus- 
Auriga and Scorpius-Ophiuchus, both at « 150 pc), the 
natural limit in angular resolution set by atmospheric tur- 
bulence will limit significantly our ability to investigate 
their properties. As a consequence, this field of research 

A. Richichi et al.: Discovery of a close companion to the young star Haro6-37 




Haro 6-37/c 


2 4 6 

Relative Time (s) 

Fig. 1. The larger figure shows a long portion of the oc- 
cultation trace, including the distant c component. These 
data are rebinned by a factor of 4 to improve the presen- 
tation. The inset is an enlargement to show the detection 
of the newly discovered companion. The occultation data 
are shown as dots and our fit by a multiple star model as 
a solid line. The horizontal bar on the bottom marks the 
portion detailed in Fig. H. 

has seen an increasingly important contribution from high 
angular resolution techniques such as lunar occultations 
(e.g. Lcinert et al. ( |l99l| ), Richichi et al. ( |1994D , Simon 
et al. (0995)) and speckle interferometry (e.g. Ghez et al. 
( |1993| ), Leinert et al. ( |1993D ). More recently, also other 
methods have been successfully employed (e.g. adaptive 
optics Roddier et al. ( |1996D , Close et al. (1998); speckle 
holography Petr et al. (1998)). Among the main results to 
be mentioned, is the fact that the frequency of binary sys- 
tems among PMS stars appears to be significantly higher 
than for main sequence stars, at least within the limita- 
tions of separation range and masses and in the number 
of stars, which are intrinsic in the currently available sur- 
veys. This is true at least for the Taurus- Auriga SFR, to 
which Haro 6-37 belongs. A detailed discussion was given 
by Simon et al. ( 1995| ). However our conclusions are still 
limited by the available statistics: in this sense, it is im- 
portant to increase the numbers of observed binary and 
multiple systems. 

With this work, we report a newly discovered com- 
panion around Haro 6-37, which we detected initially by a 
lunar occultation (LO) event, and subsequently confirmed 
by speckle interferometry observations. 







T -2s 


: A B 






Relative tiTTie(ms) 

Fig. 2. Comparison of "glitches" in relatively distant por- 
tions of the lightcurve (marked by the arrows in the two 
top curves), with the possible signature of extended emis- 
sion close to the main component (lower curve). This lat- 
ter is shown after subtraction of the fringe patterns of 
the A and B components, whose position is marked. Data 
rebinned by a factor of 4. 

2. Observations 

An occultation of Haro 6-37 was observed on November 16, 
1997 at the Calar Alto 3.5 m telescope equipped with an 
IR fast photometer. This instrument, based on an InSb de- 
tector cooled to solid nitrogen temperature, is usually de- 
voted to slit-scanning speckle interferometry, but thanks 
to its fast time response and large data storage capabili- 
ties, it has been successfully employed also to observe LO 
events, in particular of T Tauri stars (e.g. Leinert et al. 
( |1991D , Richichi et al. ( |1994| )). 

The LO data were recorded with a sampling of 1.95 ms 
in a standard wide band K filter (Aq = 2.36 /im, AA — 
0.46 /im). It is shown in Fig. fl together with our best fit. 
The fit has been obtained using a least-squares method 
which is described in Richichi et al. ( 1992| ). This approach 
is well suited to derive the parameters of a simple source 
model (such as a binary or multiple star). However, when 
an extended, geometrically complex component is present, 
a model-independent analysis is necessary. For this, we use 
an iterative method based on the Lucy's deconvolution 
algorithm, and expecially desig ned fo r LO data (the so- 
caUed CAL algorithm, Richichi ( |1989D ). 

The speckle observations were carried out also at the 
3.5 m telescope on Calar Alto shortly after the LO event 
on November 24, 1997. We used MAGIC, a 256 x 256 

pixel NICM0S3 camera (Herbst et al. ( |1993D ) in its high- 
resolution configuration at the f/45 focus, also in the K 
band. Pixel scale and orientation were determined by as- 
trometric fits to images of the Trapezium cluster, yielding 
(0'.'0710 ± 0'.'0002)/pixel and -0?28 ± 0?05 respectively. 
A sequence of 4 groups of 250 frames were recorded on 
Haro 6-37, each with an integration time of 0.8 s. These 
groups of fast integrations were alternated with a similar 

A. Richichi et al.; Discovery of a close companion to the young star Haro6-37 

Haro 6-37/c 

Haro 6-37;B 

Fig. 3. The modulus of the complex visibility of Haro 6- 
37, reconstructed from our speckle data. The frequency 
spacing and orientation of the fringe systems of the two 
companions are highlighted. The circle is a portion of the 
Fourier space with a radius of 6.2 arcsec"^. 

sequence on a nearby point -like reference star, SAO 94117. 
During data reduction this latter appeared to be itself a 
binary star, and thus not suitable for calibration. Instead, 
we used as a reference HIC 19797 (HIC = Hipparcos Input 
Catalogue), which was similarly observed in the course of 
the same observational program shortly before Haro 6-37. 
In spite of an angular distance of « 8° , the calibration by 
this star was satisfactory. 

The speckle data were processed in order to derive the 
modulus of the complex visibility (i. e. the Fourier trans- 
form of the object brightness distribution) from power 
spectrum analysis (see Fig. p|). T he phase was computed 
using the Knox-Thompson ( 1974| ) algorithm, as well as the 
bispectrum method (Lohmann et al. ( 1983 )). The modulus 
and phase were then fitted by a triple star model to derive 
the brightness ratio, separation and position angle of the 
components. Fits to different subsets of the data give an 
estimate for the standard deviation of the parameters. 

3. Results and discussion 

The occultation data shown in Fig. || reveal the presence 
of two companions to Haro 6-37. One of them is the rel- 
atively distant component firstly discovered by Moneti & 

Zinnecker ( |1991| ) by imaging, and dubbed Haro 6-37/c. Its 
projected angular distance recovered from the LO data 
is «0'.'81, but there is some uncertainty because the oc- 
cultation rate of this component appears to differ signif- 
icantly from that of the primary. This is due to the fact 
that the true separation is quite large (see below). Tak- 
ing into account the occultation geometry, it turns out 
that Haro 6-37/c was occulted by a point on the lunar 
limb 4.4 km away from the primary; this accounts for a 
significant difference in the local slope of the lunar limb, 
estimated in about 10°. At the same time, the distance 
between the two components is comparable to one-half of 
the diaphragm (6") used in our LO measurement, making 

our LO-based estimated of the brightness ratio, R«0.28, 
subject to the possible uncertainties of non-uniform detec- 
tor response across the field. In summary, the relatively 
large angular distance between these two objects makes 
it a difficult target for LO. However, Haro 6-37/c is easily 
resolved in our speckle data, as shown in Fig. pi These 
yield a true position angle PA=38?6 ± 0?1 and separation 
p=2'.'627 ± 0'.'008, with a brightness ratio R=0.35±0.01 in 
K. We note that the LO values are consistent within the 
uncertainties above mentioned. Haro 6-37/c has been rel- 
atively well studied already in the literature, and we will 
not concern ourselves further with it. 

The second, closer component is more interesting, since 
it represents a new detection. Our fit to the LO data, also 
shown in Fig. ^, yields a projected separation for the A-B 
pair pp=0'.'0857 ± O'.'OOll along PAlo=285?6 ± 0?2, with 
a brightness ratio R=0.112 ± 0.001 in K. In this case, the 
vicinity to the primary ensures a very reliable measure- 
ment. A model fit to the speckle data with the inclusion 
of this third component in the system, led to the values 
PA=180?7 ± 0?9, p=0'.'331 ± O'.'OOS and R=0.10 ± 0.02, 
also in K. It can be noted that the two independent mea- 
surements by different techniques are highly consistent, 
particularly for what concerns the separation. The bright- 
ness ratio is slightly discrepant, but we must add that the 
LO data seem to show the presence of additional extended 
emission close to the location of Haro 6-37/A. 

This was isolated from the fringe patterns of the A and 
B component and analyzed by the CAL algorithm. The 
result is consistent with an extended emission of roughly 
gaussian shape, having a projected FWHM=0'.'075 and 
traced out to a total extent of about 0'.'2. This could rep- 
resent the signature of the inner parts of the dust disk 
detected by mm-radio observations, but at the same time 
we must also point out that a few random variations of the 
background, or DC-offset in the amplification scheme, are 
also present in the data. Fig. shows some of these vari- 
ations, compared to the signal which is being tentatively 
attributed to the extended emission. Although it can be 
noted that the intensity is at least w 3 times larger for this 
latter, the order of magnitude and the typical timescales 
are comparable. Unfortunately, this prevents us to make 
a definitive conclusion. 

One consequence is that also the brightness ratio de- 
rived from the LO event must be revised. If the integrated 
intensity of suspected extended emission (be its nature 
real or not) is subtracted from that of the A component, 
the brightness ratios are 1:0.124:0.101 for the A, B and ex- 
tended components respectively. This brings the LO and 
speckle determinations for the A-B brightness ration in 
very good agreement. We also note that such extended 
emission would be extremely difficult to detect in the 
speckle observation, causing a slight, slow decrease in the 
visibility, which we cannot confirm in our data given the 
available SNR. The brightness ratio of the A-B pair would 

A. Richichi et al.; Discovery of a close companion to the young star Haro6-37 

not be affected, to a first approximation, by the neglection 
in the fit of such an extended component. 

On 9 January 1998, shortly after our LO and speckle 
observations, K-band photometry of Haro6-37 was ob- 
tained using the NIR camera IRCAM3 at the UKIRT tele- 
scope. The derived magnitude were K=(7.92 ± 0.08) mag 
and K=(8.97± 0.08) mag for the central star and the dis- 
tant companion. From the flux ratio derived above, this 
leads to K=(8.03 ± 0.09) mag for the A-l-B component, 
and K=(10.48 ± 0.09) mag for the newly detected com- 
panion. For the purpose of the discussion to follow, it is 
important to establish whether any variability is present. 
This is a common feature among T Tauri stars, although it 
is recognized that its magnitude is more important in the 
visual than in the near-IR. Hartigan et al. (1994) find that 
variability causes a mean error of 0.17 mag in the J-band 
for one component of a binary system. In our case, we can 
compare our values to the photometry obtained by Moneti 

& Zinnecker (1991) almost exactly ten years before, giving 
K=7.71mag and 8.58 mag for the A-l-B component and 
the distant companion, respectively. A decrease of about 
0.2 mag is apparent for the A+B component over this time 

The detection of a new companion is hardly surprising, 
since it is now well established that young low-mass stars 
show a high incidence of binary frequency, and particu- 
larly so in the Taurus-Auriga SFR as was noted before. 
Nevertheless this new result has some additional implica- 
tions, since it makes Haro6-37 a triple system and because 
of the presence of a circumstellar disk around it, estab- 
lished by mm-radio observations. In particular, Osterloh 
& Beckwith ( |1995| ) derive a total disk mass of 0.017 Mq, 
a characteristic temperature at 1 AU of 130 K, and speak 
of this system as an "example of a consistent picture of a 
star and a nearly standard reprocessing disk (g=0.65)". 

It is therefore interesting to look into the implications 
of the newly detected companion for the general picture 
of this system. It has been pointed out by Simon et al. 
( |l993| ), that our estimate of the ages of T Tauri stars can 
be biased up to a factor of 2 by the presence of an un- 
detected companion. Hartigan et al. (1994) have placed 
Haro6-37 and Haro6-37/c into the HR-diagram using the 
J-band magnitude as luminosity indicator. By compari- 
son with theoretical PMS tracks by D'Antona & Mazz- 

itelh (1994) and Swenson (1994), they find both stars to 

be coeval with an age of about 6 x 10 yr. The detection 
of a third component within this system could change this 
conclusion, only if it had a very red color. In fact, if we 
assume that the A-B flux ratio is constant between J and 
K, the J-band brightness of IIaro6-37/A would be over- 
estimated by about 0.10 mag only. This is not significant, 
given that Hartigan et al. (1994) estimate that variabil- 

niasses and ages, the PMS tracks are almost horizontal in 
the HR-diagram, so that if the color of Haro 6-37/B was 
significantly redder, this would imply a reduced mass and 
a larger age for the primary star, and in turn a significant 
age difference between this latter and the more distant 
companion. In any case the interesting question remains, 
if also the newly detected companion is coeval with the 
two other components. This can be answered when addi- 
tional spectra and J-band photometry become available. 
For what concerns the disk, the new companion does 
not seem to alter significantly the present understanding 
of the system. The values derived by Osterloh & Beckwith 

(1995) indicate a compact dust distribution, with an inner 
dust formation radius of only 0.013 AU, and a Ti.3mm=l 
radius of 15 AU. By comparison, the newly detected com- 
panion is at a distance of «46AU, and this should be 
considered as a minimum value for the separation in the 
absence of any orbit determination. Therefore, no real in- 
fluence on the inner dust disk is to be expected, apart from 
tidal forces or a clearing of the outer dust, effects which 
at present are beyond our observational capabilities. 

We already mentioned that the additional signal ap- 
parent in the LO data, if real, would imply an extended 
envelope with a projected FWHM=0'.'075 or about 10 AU 
and about 10% of the intensity of the primary star. The 
most natural explanation would be that this is scattered 
light from the above mentioned dust disk, and we note 
that already in other cases similar detections were ob- 

ity and interstellar extinction already produce a 0.19 mag 
uncertainty in the HR-diagram. Also the spectral type of 
the main component is essentially unaffected, because of 
the low flux ratio. However, we note that in this regime of 

tained by LO. For instance, Leinert et al. (1991) reported 
an extended emission attributed to scattered light around 
DG Tau, with FWHM of about 7 AU and 25% of the stel- 
lar brightness in K. Additionally, we note that the CAL 
analysis places this emission approximately in the mid- 
dle of the A and B components, with the two stars at 
the edges. This would suggest a circumbinary disk, but 
the picture is necessarily limited because of the projection 
effects intrinsic in the LO method. 

In conclusion, the newly detected companion does not 
alter significantly our understanding of the Haro 6-37 sys- 
tem, but it would be interesting to assess more in de- 
tail the relative spectral energy distribution of the A and 
B components, and to obtain a spectrum of this latter. 
This is at the limit of current observational possibilities, 
but would be a necessary step to determine whether also 
Haro 6-37/B is a classical T Tauri star. This would con- 
firm the result of Prato & Simon (1997) who have found 
that in 12 young binary systems with classical T Tauri be- 
haviour, both components are CTTS. Additionally, direct 
imaging at very high angular resolution of the A and B 
components could reveal the (circumbinary) disk in scat- 
tered light. Haro 6-37/B appears to be ideally suited for 
observations by the new generation of large ground-based 
imaging interferometers. 

Note added in proof. After submission of the present 
work, we became aware of a paper by G. Duchene (1999), 
in which the author quotes the discovery of the binarity of 

A. Richichi et al.; Discovery of a close companion to the young star Haro6-37 

the main component of Haro 6-37, observed with adaptive 
optics by another group (Monin et al., in preparation). 
The reported result, obtained just one month after our 
detection, is in very good agreement with the values given 
in the present work for what concerns angular separation 
and position angle. Some disagreement with our deter- 
mination seems to exist for the brightness ratio, which 
Duchcne reports as AK=1.57±0.05. 

Acknowledgements. This research has made use of the Simbad 
database, operated at CDS, Strasbourg (France). We thank P. 
Kalas for obtaining the UKIRT photometry. 


Close L.M., Dutrey, A., Roddier F., et al. 1998, ApJ 499, 883 

D'Antona F., MazziteUi I. 1994, ApJS 90, 467 

Duchene G. 1999, A&A, 341, 547 

Chez A. 1998, ASP conference series 148, 260 

Chez A., Neugebauer C, Matthews K. 1993, AJ, 106, 2005 

Hartigan P., Strom K.M., Strom S.E. 1994, ApJ 427, 961 

Hartmann L., 1998, Accretion Processes in Star Formation, 

Cambridge Astrophysics Series 32 
Herbst T.M., Birk C, Beckwith S.V.W., et al. 1993, Proc. SPIE 

1946, Fowler, A.M. (ed.), p. 605 
Knox K.T., Thompson B.J. 1974, ApJ 193, L45 
Leinert Ch., Haas M., Richichi A., Zinnecker H., Mundt R. 

1991, A&A 250, 407 
Leinert Ch., Zinnecker H., Weitzel N., et al. 1993, A&A 278, 

Lohmann A.W., Weigelt C, Wirnitzer B. 1983, Appl. Opt., 

22, 4028 
Mathieu, R.D., 1994, ARA&A 32, 465 
Moneti A., Zinnecker H. 1991, A&A 242, 428 
Osterloh M., Beckwith S.V.W. 1995, ApJ 439, 288 
Petr M.G., Beckwith S.V.W., Coude du Forcsto V., Richichi 

A., McCaughrean M.J. 1998, ApJ 500, 825 
Prato L., Simon M. 1997, ApJ 474, 455 
Richichi A. 1989, A&A 226, 366 

Richichi A., Lisi F., Di Giacomo A. 1992, A&A 254, 149 
Richichi, A., Leinert Ch., Jameson R., Zinnecker H. 1994, A&A 

287, 145 
Roddier C, Roddier F., Northcott M.J., Graves J.E., Jim K. 

1996, ApJ 463, 326 
Simon M., Ghez A.M., Leinert Ch. 1993, ApJL 408, 33 
Simon M., Ghez A.M., Leinert Ch., et al. 1995, ApJ 443, 625 
Swenson F.J., Faulkner J., Rogers F.J., Iglesias C.A. 1994, ApJ 

425, 286